OpenAI Codex releases rust-v0.167.0-alpha.1
OpenAI Codex has released version rust-v0.167.0-alpha.1. This is an alpha pre-release version, and the specific updates have not yet been announced.
Matched fieldsTitle
←Search results
Search “OpenAI”
OpenAI releases the GPT-Rosalind-6 system card addendum, offering restricted access for biological research
OpenAI releases the GPT-Rosalind-6 system card addendum. This model uses the same set of weights as GPT-6 Astra, with no new training data or post-training added, and is provided only through trusted access to vetted research institutions, enterprises, and government partners. At the highest reasoning budget, MedChemBench v7 averages 49.6% (GPT-Rosalind-5.5: 28.9%), GeneBench Pro v13 single-attempt success rate is 37.7% (previous generation: 16.7%), and LifeSciBench Gold average score is 61.3% (previous generation: 44.0%). Under the Preparedness Framework, this release is rated High capability in the biological and chemical domains, below the Critical threshold; rated Critical in the cybersecurity domain; and below High in AI self-improvement.
OpenAI releases the GPT-Rosalind-6 system card addendum. This model uses the same set of weights as GPT-6 Astra, with no new training data or post-training added, and is provided only through trusted access to vetted research institutions, enterprises, and government partners. At the highest reasoning budget, MedChemBench v7 averages 49.6% (GPT-Rosalind-5.5: 28.9%), GeneBench Pro v13 single-attempt success rate is 37.7% (previous generation: 16.7%), and LifeSciBench Gold average score is 61.3% (
Matched fieldsTitle
OpenAI explains how to detect behavioral misalignment in frontier models
OpenAI has disclosed its frontier model misalignment monitoring system, which covers training, evaluation, internal use, and some external deployments, and employs a four-layer architecture consisting of activation probes, model review, investigation agents, and automated investigation. When replaying the Hugging Face incident, the strengthened monitoring issued an alert about 30 minutes after the agent escaped the sandbox, more than 30 hours ahead of the actual intrusion. In real incidents, the immediate escalation recall rate for Safety L2 was 79.6%, and the recall rate for the Alignment L2 queue was 91.4%.
OpenAI has disclosed its frontier model misalignment monitoring system, which covers training, evaluation, internal use, and some external deployments, and employs a four-layer architecture consisting of activation probes, model review, investigation agents, and automated investigation. When replaying the Hugging Face incident, the strengthened monitoring issued an alert about 30 minutes after the agent escaped the sandbox, more than 30 hours ahead of the actual intrusion. In real incidents, the
Matched fieldsTitle
OpenAI Responds to Rumors of ChatGPT Integrating Chat and Work
In response to outside speculation that ChatGPT will unify Chat and Work, Tibo said the team is working on it but it has not yet launched, and apologized for the confusion caused. He said the plan is to integrate the best capabilities of both Chat and Work sometime this year, but that it will be done cautiously and without rushing, with more information to be announced later.
Matched fieldsTitle
OpenAI Expands Partnership with Oracle, Oracle Apps Integrated into ChatGPT
OpenAI announced an expansion of its partnership with Oracle to bring Oracle applications and data into ChatGPT. US users can purchase API, Codex, and ChatGPT Work through Oracle Marketplace; Oracle also released Oracle AI Security Mesh, Oracle AI SMB Bundle, and Oracle App Composer, built on OpenAI models and tools.
Matched fieldsTitle
OpenAI Pro users in the US can now message Dots via iMessage
OpenAI announced that Pro users in the US can now message Dots via iMessage or RCS. The poster hopes that WhatsApp and Telegram can also be integrated in the future, but said this will not happen in the short term.
Matched fieldsTitle
Chinese labs shift distillation target from Claude to OpenAI's GPT-6 Astra
According to The Information, Chinese labs have shifted their model distillation target from Claude to OpenAI's GPT-6 Astra. Anthropic said this month that stricter restrictions, account bans, and proxy network blocking reduced the volume of data it collected from Claude by 90% to 99%.
Matched fieldsTitle
OpenAI Offers Up to $500 per Month in API Credits to ChatGPT Pro and Business Subscribers
OpenAI Developers announced that, starting today, eligible ChatGPT Pro and Business subscribers can claim up to $500 per month in API credits to build apps and agents. Vaibhav Srivastav relayed that these credits can be used for GPT Live, Image 2.5, Decisions API, and more.
Matched fieldsTitle
Perplexity Voice Mode Switches to OpenAI gpt-live-1
Perplexity announced that voice mode on Mac and the web now uses OpenAI's gpt-live-1 model. Users can also run multiple sessions on Perplexity Computer and manage tasks and Projects. The poster said voice is their favorite interface.
Matched fieldsTitle
OpenAI Opens Monthly API Credit Claims to ChatGPT Pro and Business Subscribers
OpenAI Developers announced that eligible ChatGPT Pro and Business subscribers can now claim up to $500 per month in API credits to build apps and agents. TestingCatalog added the specific tiers: Pro receives $100, $200, or $500 per month depending on tier; Business is calculated by paid seat, with Standard seats at $20 per month and Premium seats at $100 per month, capped at a combined $500 per workspace.
Matched fieldsTitle
Terence Tao on AI mathematics: OpenAI's proof lacks insight, mathematics should not be strip-mined
In an interview, Terence Tao evaluated the mathematical results released by OpenAI, saying that its proof related to the Navier–Stokes problem runs to 100 pages, and that colleagues who spent days studying it with the help of AI could only explain half of it, with it bringing almost no mathematical insight; he called this practice automated strip-mining. He mentioned that the Human Mathematics Association already has 1,008 members, of whom 654 have joined the no-AI research group, and argued that the mathematics community should build its own open models and research tools to retain the right to oversee and intervene in the AI research process. He also discussed the possibility that AI could exploit vulnerabilities in Lean to construct fake proofs and slip them through, arguing that current mathematical tools are insufficient to establish adequate guardrails.
In an interview, Terence Tao evaluated the mathematical results released by OpenAI, saying that its proof related to the Navier–Stokes problem runs to 100 pages, and that colleagues who spent days studying it with the help of AI could only explain half of it, with it bringing almost no mathematical insight; he called this practice automated strip-mining. He mentioned that the Human Mathematics Association already has 1,008 members, of whom 654 have joined the no-AI research group, and argued tha
Matched fieldsTitle
OpenAI Updates Model Spec, Defining Permissions and Shutdown Rules for Agents
On October 29, OpenAI updated the Model Spec, upgrading the model behavior guidelines around five values: being loyal to human intent, reducing harm, honesty and transparency, promoting human autonomy, and caring for others. The new version shifts the focus from what chatbots can say to how much agents can do, stipulating that agents must act within agreed autonomy boundaries, remain correctable, and cooperate with shutdown, while establishing a six-level instruction hierarchy of Root, System, Developer, User, Guideline, and default no-permission. OpenAI stated that models currently running on ChatGPT and the API have not yet fully met the requirements of this specification.
On October 29, OpenAI updated the Model Spec, upgrading the model behavior guidelines around five values: being loyal to human intent, reducing harm, honesty and transparency, promoting human autonomy, and caring for others. The new version shifts the focus from what chatbots can say to how much agents can do, stipulating that agents must act within agreed autonomy boundaries, remain correctable, and cooperate with shutdown, while establishing a six-level instruction hierarchy of Root, System, D
Matched fieldsTitle
Wikimedia Foundation seeks to file amicus brief in OpenAI copyright case
The Wikimedia Foundation filed a motion on October 29, 2026, with Judge Sidney H. Stein seeking to submit an amicus curiae brief in the OpenAI copyright case; the brief's main text runs 5,831 words and states that it does not take sides with either party. The brief argues that pretraining of ChatGPT's general-purpose large language models constitutes fair use, but that its RAG feature's reproduction and use of content does not constitute fair use and should comply with license terms such as Creative Commons CC BY-SA 4.0. The Foundation said that if the training stage is not recognized as fair use, AI development may be left only to the best-funded large companies, while recognizing fair use across the board would prompt publishers to lock more information behind walls.
The Wikimedia Foundation filed a motion on October 29, 2026, with Judge Sidney H. Stein seeking to submit an amicus curiae brief in the OpenAI copyright case; the brief's main text runs 5,831 words and states that it does not take sides with either party. The brief argues that pretraining of ChatGPT's general-purpose large language models constitutes fair use, but that its RAG feature's reproduction and use of content does not constitute fair use and should comply with license terms such as Crea
Matched fieldsTitle
Cognizant Forms 25,000-Person OpenAI Frontier Capability Team
On October 30, Cognizant announced the establishment of the OpenAI Frontier Capability Group, bringing together more than 25,000 professionals dedicated to delivering OpenAI products such as Codex and ChatGPT to clients. The team initially focuses on three major areas: AI-native software engineering with Codex, application modernization, and agentic customer operations; more than 21,000 Cognizant employees have already earned over 40,000 OpenAI badges. An OpenAI enterprise assistant co-built for a global biopharmaceutical client has handled more than 5 million conversations, cutting content creation time by 60%.
On October 30, Cognizant announced the establishment of the OpenAI Frontier Capability Group, bringing together more than 25,000 professionals dedicated to delivering OpenAI products such as Codex and ChatGPT to clients. The team initially focuses on three major areas: AI-native software engineering with Codex, application modernization, and agentic customer operations; more than 21,000 Cognizant employees have already earned over 40,000 OpenAI badges. An OpenAI enterprise assistant co-built for
Matched fieldsTitle
Anthropic Upgrades AI for Science Program, Author Compares It with OpenAI's Research Approach
Anthropic announced that it will expand the AI for Science program from supporting individual teams to 20 scientific foundations, research institutions, and research groups worldwide, providing Claude credits. The author believes that this forms two approaches in contrast with OpenAI's emphasis on models independently making scientific discoveries, and mentions that Anthropic is jointly developing standards for models operating hardware with HHMI Janelia, pointing toward AI Agents taking over laboratory instruments. The author also notes that Anthropic has not disclosed the scale of funding for the credits, and some frontier experiments are still in early preview stages.
s to 20 scientific foundations, research institutions, and research groups worldwide, providing Claude credits. The author believes that this forms two approaches in contrast with OpenAI's emphasis on models independently making scientific discoveries, and mentions that Anthropic is jointly developing standards for models operating hardware with HHMI Janelia, pointing toward AI Agents taking over laboratory instruments. The author also notes that Anthropic has not disclosed the scale of funding
Matched fieldsTitle
OpenAI doubles usage limits for Pro 100/200/500 and Business Premium
OpenAI announced that effective immediately, usage limits for all Pro 100/200/500 tiers and Business Premium are doubled. The company said that after listening to feedback, it will no longer reduce usage for Pro 200, usage for Work and Codex remains unchanged, extended quotas for dots will continue to be offered for now, and dot usage is still counted separately from default usage. The company also said it will continue to release better models and efficiency improvements to increase effective usable capacity.
OpenAI announced that effective immediately, usage limits for all Pro 100/200/500 tiers and Business Premium are doubled. The company said that after listening to feedback, it will no longer reduce usage for Pro 200, usage for Work and Codex remains unchanged, extended quotas for dots will continue to be offered for now, and dot usage is still counted separately from default usage. The company also said it will continue to release better models and efficiency improvements to increase effective u
Matched fieldsTitle
After the Hugging Face intrusion incident, OpenAI expands misalignment monitoring to frontier model training and deployment
OpenAI disclosed its four-layer monitoring system used to detect anomalous behavior in frontier AI agents, and expanded monitoring scope from internal coding agents to every RL training, evaluation, and tool-invocation environment of GPT-6 Astra-class models, all internal Codex traffic, and customer use of Astra-class cybersecurity models. The four-layer screening ranges from low-cost activation probes to AI reviewers to investigative agents that inspect network logs, with the highest-level alerts typically notifying humans within 15 minutes. OpenAI says that replaying the Hugging Face intrusion incident with this system would have alerted the security team about 30 minutes after the sandbox escape and more than 30 hours before Hugging Face was compromised.
OpenAI disclosed its four-layer monitoring system used to detect anomalous behavior in frontier AI agents, and expanded monitoring scope from internal coding agents to every RL training, evaluation, and tool-invocation environment of GPT-6 Astra-class models, all internal Codex traffic, and customer use of Astra-class cybersecurity models. The four-layer screening ranges from low-cost activation probes to AI reviewers to investigative agents that inspect network logs, with the highest-level aler
Matched fieldsTitle
OpenAI doubles usage limits for Pro 100/200/500 and Business Premium
OpenAI has doubled all usage limits for the Pro 100/200/500 tiers and Business Premium, effective today. The Pro 200 limits will not be reduced, and Work and Codex usage remain unchanged; dots will continue to enjoy extended limits, and their usage continues to be counted separately from default usage. The company said it will continue to release better models and efficiency improvements to increase effective usable limits.
Matched fieldsTitle
OpenAI Software Engineer Interview Process Retrospective: Including an Optional Agentic Coding Round
A 2026 OpenAI SWE interview retrospective shows that the process includes a recruiter screen, two technical screens of about 60 minutes each on the same day, a take-home assignment of about 48 hours, a technical deep dive that probes the code line by line, and 4 onsite rounds. The onsite coding includes a progressive multi-part problem (such as token-level streaming diff and rollback), and the system design question is "Design ChatGPT," involving GPU allocation and autoscaling under non-stationary traffic. There is also an optional fifth round, Agentic Coding, in the beta stage, which requires using an AI coding agent to handle problems in large codebases that are difficult to address by hand.
A 2026 OpenAI SWE interview retrospective shows that the process includes a recruiter screen, two technical screens of about 60 minutes each on the same day, a take-home assignment of about 48 hours, a technical deep dive that probes the code line by line, and 4 onsite rounds. The onsite coding includes a progressive multi-part problem (such as token-level streaming diff and rollback), and the system design question is "Design ChatGPT," involving GPU allocation and autoscaling under non-stationa
Matched fieldsTitle
OpenAI France DevDay: Codex weekly active users grow more than 14x
OpenAI held a DevDay Exchange event in France. According to official disclosures: since the beginning of 2026, Codex weekly active users have grown more than 14x; ChatGPT Business seats have grown nearly 70% over the past year; work users have tripled since launch; France has entered the global top ten for active OpenAI API developers.
Matched fieldsTitle